This is our second competing continuation for a project begun in 1983. Work during the most recent funding period (1987-present) has resulted in 27 publications that lay out the steps that have established an animal model of amnesia in the monkey and in identifying the anatomical components of the medial temporal lobe memory system. These developments allow us to now address a fundamental issue concerning memory function, that is, how individual structures of this system contribute to memory. The present proposal involves detailed behavioral and anatomic analyses of a group of normal monkeys and seven operated groups of monkeys including those with damage to components of the medial temporal lobe memory system and to area TE. Two major long-standing issues will be addressed: (1) Short-term Memory. Despite its theoretical importance, and its sparing in human amnesia, short-term memory has rarely been assessed in monkeys or other animals. We will determine whether damage to the medial temporal lobe spares short-term memory. A key aspect of this component of the project is to ask whether the pattern of performance on tests of short-term memory following lesions that involve the perirhinal and parahippocampal cortex resembles more closely the pattern associated with lesions of the hippocampus or the pattern associated with TE lesions. (2) Specialization of Function Within the Medial Temporal Lobe. We will determine if the components of the medial temporal lobe memory system have specialized memory functions by determining if lesions of these structures result in disproportionate deficits in any of two domains of memory relative to recognition memory: objective-reward association and spatial memory. We will also reference the effects of lesions in the medial temporal lobe, which are predicted to affect memory without affecting visual processing, to the effects of damage in an area where we expect to find a visual processing deficit. Thus, we will also determine the effect on these domains of lesions to area TE. This work is advantaged by new facts about the neuroanatomy of the primate temporal lobe that make differential predictions about the effects of specific lesions on visual and spatial memory task performance. In addition, we will compare the effects of circumscribed hippocampal and amygdala lesions to the effects of larger lesions that include adjacent cortex. A finding that hippocampal damage effects spatial memory to the same degree as do larger lesions would suggest that the hippocampus has a special role in spatial memory.